Disruption and the New Financial World Order (I)

How Advances in IT and AI will undermine--and eventually displace--the traditional investment management model

The historical evolution of the investment industry is one that has largely been shaped by constraints rather than opportunity. That may be about to change. Drastically.

The first widely accepted attempts at introducing rigorous formalism into investment decision making in the middle of the twentieth century, with the mainstreaming of insights from the likes of Graham & Dodd and Markowitz, coincided with a rapid growth of popular interest in financial markets and the emergence of corporate pension and other institutional funds as dominant players in the landscape. Ever growing demand for financial securities from the general public, and the new generation of intermediaries which emerged to cater to it, helped drive spectacular growth both in the size and influence of the industry, such that by the mid 60s and 70's the financial markets (especially for equities) had become central to policy-making for corporations and governments in developed markets, particularly with regard to old age pensions. It is easy to overlook their impact, but these phenomena--expansion and the related trend towards intermediation--had a profound influence in shaping the subsequent architecture of the investment management industry as we have come to recognise it today.

The tortuous road to today

With rapid growth, the industry quickly hit the limits of the then prevailing business model of individualised advisory. With the resources available--a handful of advisors, analysts and portfolio managers--it was simply impossible for investment managers to profitably tailor their services to individual end-investors in a rapidly growing market. The industry responded by aggressively marketing a collectivist approach to the public, cementing the ascendancy of fund structures and the institutional model of investment management that has prevailed ever since. This scalability problem in its service model remains a fundamental one in the industry's architecture, though it is rarely recognised as such. Indeed, institutionalisation of financial markets, and the professionalism it is presumed to facilitate, is often lauded as beneficial to the global financial system even though research indicates that, among other agency problems, it encourages short-termism, herding and pro-cyclicality in investment decision-making, exacerbating asset price volatility and bubbles.

Coincidentally, starting from the late 50s, the two predominant paradigms for investment decision making entered an interesting and still evolving dynamic. On the one hand was the formalism introduced by Markowitz's Modern Portfolio Theory (MPT) and its spinoffs, which emphasised the notion of market efficiency and firmly established a standardised concept of performance that incorporates both the return and risk dimensions. On the other hand were the fundamentalists, followers of Graham & Dodd, led by such industry luminaries as Berkshire's Warren Buffet and Fidelity's Peter Lynch, who argue that careful analysis could uncover a select group of securities with outstanding long term performance well beyond what the broader markets could deliver.

Throughout the 1970 and 80's, these sometimes dueling concepts led to the proliferation of smaller 'specialist' managers who marketed their ability to outperform the broad market indices indicated by MPT, typically via fundamental research, charging enormous fees for the privilege. Nevertheless as a testament to the depth of MPT's influence, sponsors still attempted to conform to the portfolio approach it advocated, adopting the "manager of managers" (or fund of funds) model, ostensibly to harness the best of both worlds, but introducing yet another layer of intermediation and fees. However, while some specialist 'active managers' initially appeared to enjoy substantial investment success and the financial rewards that came with it, diminishing returns had set in by the late 80s and early 90s and, alongside ever more stringent measures of performance, led to increased scrutiny of their claims.

With increasing doubts about active management, a new paradigm more in tune with MPT and championed by the likes of Vanguard's John Bogle gradually began to predominate. Already emergent in the mid 70s, practitioners of the so called 'passive management' approach--which limited itself to efficiently replicating broader indices under the assumption that attempts to outperform these were ultimately futile--began to lure supra-managers and their sponsors with the promise of more consistent performance at significantly reduced costs. This was aided in no small measure by theoretical and practical refinements to MPT, including a more meticulous distinction of investment styles and asset classes alongside the introduction of the concept performance 'alpha'--attributable to the manager's inherent skill--and 'beta' to measure the performance impact of his (possibly inadvertent) market exposure; all of which accelerated with the major market crises of the 90s and 2000s, resulting in significant levels of concentration in both managers and styles in the passive space.

Nevertheless, so strongly ingrained is this notion that investors have never really given up on outperforming "Mr Market". Today, the trend of ever more specialised refinements of the indexing paradigm via the so-called 'smart indexing', which attempts to harvest increasingly esoteric features of securities and asset classes for more effective portfolio construction, continues to captivate the industry. Notable in all these changes was the decisive influence of MPT which, despite its elegant formulations, then as now, had scant empirical support for its underlying assumptions. Moreover, despite this evolution, this "funnel model", i.e. collectivisation of beneficiaries investments into the various varieties of fund structures--remains unchallenged as the industry's prevalent architectural paradigm which, driven more by industry contingencies and service-provider convenience than by client needs, also helped perpetuate the blind spot around the scalability problem, whose illumination may have profound implications for industry evolution going forward.

Technology takes centre stage

Financial markets rarely lag any other facet of the economy in technology adoption; and many of its key aspects, including the architecture and operations of securities exchanges and other market infrastructure, as well as the dissemination of investment data and products, have been thoroughly transformed by technology, a trend which has accelerated since the 1980s, closely mirroring the rapid advancement of information technology during this period. However, it was not until the last decade or so that a confluence of factors, mostly centered around technology, have begun to create a viable pathway to addressing the scalability problem in the investment management service model. We outline some of these factors that will likely converge to transform the architecture and perhaps the very nature of investment management industry in coming years.

Enter, the data deluge

The mass adoption of smartphones over the past decade was yet another watershed moment in the history of information technology, arguably on par with the proliferation of the personal computer in the 80s and the mainstreaming of the internet at the tail end of the last millennium in its impact. On their own, the resultant explosion-cum-miniaturisation of computational power and the momentous upsurge in our information dissemination capabilities was already pivotal, but Lady Technology was not done yet. It still deigned to confer on humanity, the distributed computing paradigm and its wonderful offspring, the cloud; the distributed ledger and cryptographic technologies that encompass blockchain; the magic of interconnected devices; and a veritable cornucopia of algorithms, frameworks and concepts that underpin today's incarnation of artificial intelligence and machine learning, among other benefices; all in roughly the same time-frame.

Consequently, not only are we confronted with an absolute Cambrian explosion of the amount of data we are now able to capture about our lives and interactions--economic and otherwise--we now have, for the first time in our existence as a species, means to conceivably grapple with, and make sense of these. It may be well nigh impossible to quantify the effect of these trends on the arc of human history but there is no question that many aspects of our lives are being, and will continue to be, transformed by these events, finance being no exception. In the current context, the reduction of coordination costs facilitated by the convergence of these technologies set the stage for unleashing powerful network effects across the entire financial services ecosystem.

A breakthrough in the service model

From the foregoing, it is fair to say that data deluge and AI has finally opened new avenues for mass-personalisation. On the one hand, data streams from end-users' economic and financial interactions can provide insight into their true preferences and a more holistic context for their needs, opening the door for cost effective customisation of products and solutions. On the other hand, the factors outlined in this discussion not only create new frameworks for better comprehending markets as a collection of opportunity sets, but a pathway for cost-effectively harnessing this knowledge for client benefit in an individualised manner. Clearly, this requires reconfiguration of existing processes and re-imagination of many facets of the financial services value chain, from delivery platforms to performance evaluation models. However, even today, the fintech space is replete with implementation of these ideas even if they have yet to achieve mainstream status.

A popular example, for starters, would be the booming US robo-advisors like Betterment and Wealthfront. Attempts to maintain alignment with the existing industry architecture probably accounts for their heavy inclination towards MPT, but the planning tools and high levels of customisation their platforms offer already hint at one future industry mass-market model.

However, a more pertinent--and spectacular--example would be Ant Financial, the financial services arm of the Chinese tech giant, Alibaba, which became the largest money market fund in the world in 2017 by leapfrogging decades-old global rivals in the space of a mere three years. It did this by leveraging its multiple consumer facing platforms into a ecosystem for cost-effectively addressing multiple financial pain points in such areas as payments, insurance, wealth management, credit etc. In response to a regulatory clampdown, it then pivoted to creating a platform for traditional wealth managers to utilise Ant's AI-enabled consumer interface to offer user-tailored financial services. So successful has this deployment of a classic 'flywheel' approach been, that as of 2019 ~60% the traditional wealth managers in China were reportedly using their platform to reach customers, with some of the enlisted reporting triple digit annual growths in transactions volumes as a result.       

It is sometimes argued that the investors will always need human advisors eventually, especially to provide reassurance in times of financial turmoil, which would presumably curtail the widespread adoption of the automated investing paradigm. However, the experience of Ant Financial challenges this view, which moreover misses the point in that not only is robo-advisory not really a substitute (it is probably the only viable option for mass distribution of financial services of any real sophistication) it ultimately represents a take on the industry architecture different from that which this view is derived, as this essay tries to highlight.

In the second part of this essay we will explore how these trends will conspire to fundamentally alter the investment industry.

to be continued...